Redis SAs - Memcached Sizing Calculator

A minimal tool to measure dataset size and throughput (ops/sec) of a running Memcached instance.

This is useful for:

• Estimating memory usage of your current keyspace
• Measuring read/write throughput based on real-time stats

🧰 What's Inside

.
├── memcached_sizing_calculator.py  # Main tool: measures memory and ops/sec
├── load_generator.py               # Optional helper: writes dummy data for testing
└── requirements.txt                # Python dependencies

✅ Features

• 📊 Calculates total dataset size in MB
• ⚡ Estimates ops/sec by sampling cmd_get and cmd_set over time
• 📎 Uses official python-memcached client (no deprecated telnetlib)

🚀 Getting Started

1. Install Dependencies

python3.12 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

2. Run Memcached

Locally:

memcached -p 11211 -m 512 -vv

Or via Docker:

docker run -d --name memcached -p 11211:11211 memcached

3. Populate Some Data (Optional)

Use the included generator to create keys/values:

python load_generator.py

This is useful if your Memcached instance is empty and you want to simulate usage.

4. Run the Calculator

python memcached_sizing_calculator.py

This will:

• Connect to Memcached
• Wait 60 seconds (default)
• Print dataset size and ops/sec

✍️ Sample Output

--- Memcached Metrics ---
Dataset Size: 16.42 MB
Ops/sec (get + set): 2023.40
 - cmd_get delta: 10200
 - cmd_set delta: 2000
 - Total keys: 10000
 - Memory cap: 64.00 MB

⚙️ Config

You can tweak host, port, and interval inside memcached_sizing_calculator.py:

HOST = "127.0.0.1"
PORT = 11211
INTERVAL = 60  # in seconds

📄 License

MIT